ABSTRACT
Modelling time series data and the temporal processes that generate them is of paramount importance in environmental epidemiology where we find several areas of application. These include modelling underlying patterns in exposures, for example to air pollution, as well as temporal patterns in health outcomes. Predictions may be made both within the time frame of the given data and in the future, the latter of which it is known as forecasting. An example of forecasting in environmental processes is where urban areas produce 24 hour ahead forecasts of air pollution levels (Dou, Le, & Zidek, 2012). This chapter contains a background to the study of temporal processes, which replaces space, the subject of Chapter 9, as the domain of interest whilst drawing on many of the concepts from the previous chapter.
